ABSTRACT
Virulent Rhodococcus equi causes pyogranulomatous bronchopneumonia in foals. The route of infection of foals has been considered to be inhalation of aerosolized bacteria from soil that is contaminated with equine feces. Thus, disease caused by R. equi has been regarded as an opportunistic infection of environmental origin and not a contagious disease. In this study, we report the exhalation of virulent R. equi from the respiratory tract of naturally infected foals. A handheld air-monitoring system was used to recover virulent R. equi from the exhaled breath of foals, and the concentration of virulent R. equi organisms in exhaled air was compared to the concentration in environmental air samples taken from the holding pens and lane areas on farms. R. equi strains carrying the vapA gene of the virulence plasmid were detected by using colony blotting and DNA hybridization techniques in cultures of exhaled air from 67% (37/55) of foals tested. The concentration of virulent R. equi organisms in exhaled air from foals was significantly higher than that in environmental air (P < 0.001). There were no significant differences in the median concentrations of virulent R. equi bacteria exhaled by clinically healthy or diseased foals. The high concentrations of virulent R. equi bacteria in exhaled air suggested that aerosol transmission between foals is possible and may have a significant impact on the prevalence of R. equi pneumonia on farms. The air sampling technique described is potentially useful as a noninvasive method for the detection and quantification of virulent R. equi in the respiratory tract of foals.
Rhodococcus equi is a gram-positive coccobacillus found in the soil and the feces of grazing herbivores (3). Virulent strains of R. equi cause bronchopneumonia in horses, primarily affecting foals between 1 and 4 months of age (5, 10, 19, 29). Virulent R. equi contains an 85- to 90-kb plasmid that carries genes for virulence-associated proteins (Vaps) (24, 25). The virulence plasmid and these Vaps, particularly the VapA cell surface protein, are considered the factors that enable these strains to induce pneumonia in foals (6, 13). Recent vap mutagenesis studies have found that reversion to virulence occurred on the insertion of vapA into the virulence plasmid but not after insertion of the vapC, -D, -E, or -F gene (13). This finding highlights the importance of vapA in virulence and vindicates its use as a fundamental criterion for the identification of virulent R. equi in diagnostic assays in horses and epidemiological studies over the last decade (5, 15, 22, 23). The accepted pathogenesis of R. equi pneumonia is that virulent R. equi bacteria in the soil are aerosolized, inhaled, and phagocytosed by alveolar macrophages, in which they multiply and cause pneumonia. This hypothesis has been accepted by investigators because the organism is a soil saprophyte and is detectable as part of the healthy gastrointestinal flora of horses. Growth of R. equi has been shown to be enhanced (up to 10,000-fold) in the presence of volatile fatty acids common in equine feces (3, 11). Intestinal carriage has been demonstrated in adults, but foals up to 12 weeks of age and particularly those with clinical disease have been found to have considerably higher concentrations of R. equi organisms in their feces than adults. Thus, feces from foals are thought to be the main source of soil contamination (23, 26, 27). It has been suggested that as the concentration and proportion of the virulent organisms in the environmental R. equi population increase, there is an associated increased risk of inhalation of virulent R. equi from the soil (22), and thus, an increase in the prevalence of disease caused by R. equi infection (23, 27). Recent studies have also shown that higher concentrations of airborne virulent R. equi and high proportions of virulent R. equi among airborne R. equi populations are associated with an increased prevalence of R. equi pneumonia (15).
Areas frequented by foals and environmental conditions that favor multiplication in the soil and aerosolization of R. equi from the soil, such as poor pasture cover or dry or sandy soils, are considered higher risks for foals (15).
There are many reasons why R. equi pneumonia has not been considered a contagious respiratory disease. The intracellular habitat of the organism in the alveolar macrophages of infected foals, the sometimes problematic recovery of R. equi from tracheal lavage fluid samples from foals with clinical R. equi pneumonia, and the production of abscesses in the lungs of infected foals have suggested that direct animal-to-animal transmission is unlikely (2, 8, 9, 14). However, if foal-to-foal transmission by the aerosol route were possible, virulent R. equi might be acquired more readily within a closely confined group of foals than by the inhalation of aerosolized bacteria from soil. This could explain the variable efficacy of environmental management strategies, such as irrigation of pastures, pens, and laneways and collection of feces, in reducing the R. equi disease burden in herds.
The aim of this study was, first, to establish whether virulent R. equi could be detected in the exhaled air of foals with or without R. equi pneumonia. If virulent R. equi was detected in exhaled air samples, then a second aim was to compare the concentrations of airborne virulent R. equi organisms in high-risk environmental areas with those detected in the exhaled air of foals. In this way, the possibility that virulent R. equi would be spread between foals and its significance in comparison with the currently accepted horse-environment cycle of infection could be evaluated.
MATERIALS AND METHODS
Collection and analysis of samples.Air samples were collected by using a portable air-monitoring system (M Air T; Millipore) onto ceftazidime-novobiocin agar modified by the addition of cycloheximide (40 μg/ml) as an antifungal agent (mCAZ-NB) (15, 16, 18). Cultures were blotted as described previously (16, 17), with virulent R. equi (isolate 7) and avirulent R. equi (isolate 128) organisms as controls and a negative control (Bacillus fumarioli) incorporated into each blot. The blots were probed with a 32P-labeled PCR product amplified from the virulent R. equi vapA gene to identify virulent R. equi colonies, as described previously (16). Briefly, a single 500-liter environmental air sample was taken in the holding pens (small outdoor fenced areas used to confine horses prior to procedures such as breeding of mares, veterinary treatment, and farriery) and lanes along which horses were moved to and from pens and paddocks. Exhaled air samples from foals were collected by holding the air-monitoring device to the muzzle of a manually restrained foal. A single 100- or 250-liter sample was collected from each foal. The procedure took between 1 and 2 min, depending on the volume of air sampled. The sieve was disinfected with an isopropanol wipe (Isowipe; Kimberly-Clark, Milson's Point, NSW, Australia) before the collection of each air sample. The effectiveness of this disinfection method was tested prior to sampling. Ten microliters of broth-cultured R. equi (∼107 colony-forming units [CFU]/ml) was used to inoculate four different areas of the sieve, and 1,000-liter air samples were collected onto mCAZ-NB agar. The sieve was disinfected, and the sampling was then repeated. The inoculation of the sieve, disinfection, and air sample collection were repeated another three times. The agar plates were incubated for 48 h at 37°C and examined for colonies of R. equi. R. equi was recovered from air samples on the regions of the agar plates that corresponded to the areas of the sieve that were inoculated but only from the samples collected prior to disinfection in each experiment. No R. equi was recovered from samples taken after disinfection with the isopropanol wipes, indicating that this method of disinfection was adequate for decontamination of the air sampler between collections of samples in the field.
Farms and foals.Exhaled air samples were collected from 55 foals on eight Thoroughbred farms in Victoria and New South Wales, Australia, during the 2000 and 2001 breeding seasons. Samples were collected from foals at a single point in time after diagnosis or at the time of ultrasonographic lung examination. Environmental air samples were collected monthly between November 2000 and February 2001 from the lanes and holding pens on six Thoroughbred farms. All six farms from which environmental air samples were collected reported cases of R. equi pneumonia during the sampling period. On three of these six farms, exhaled air samples were collected from foals with R. equi pneumonia.
There was no formal, randomized sampling frame used to determine which foals were sampled. Initially, only foals with R. equi pneumonia were selected to test the hypothesis that virulent R. equi could be detected in exhaled air. The foals from which exhaled air was collected ranged in age from 25 to 90 days. Of these, 35 foals (64%) were from two farms that had a prevalence of R. equi pneumonia of over 9% during the study period. Forty-five of the 55 foals were being treated for R. equi pneumonia when they were sampled. All 45 foals treated for R. equi pneumonia had clinical signs of pneumonia and at least one other diagnostic feature of R. equi pneumonia: specifically, ultrasonographically detectable pulmonary abscesses (n = 36), leukocytosis with neutrophilia and fibrinogenemia (n = 37), and/or an R. equi-positive culture from a tracheal lavage fluid sample (n = 10). Eleven foals on one farm were sampled while they were being restrained for routine thoracic ultrasonographic examination. Seven of these foals were considered healthy, with no ultrasonographically detectable lung lesions suggestive of R. equi pneumonia, while one had been diagnosed with R. equi pneumonia prior to sampling. The remaining three foals had ultrasonographically detectable lung abscesses and were diagnosed with R. equi pneumonia, despite the absence of overt signs of pneumonia.
Data and statistical analysis.Median concentrations of virulent R. equi organisms in the environmental and the exhaled air samples and the median concentrations of virulent R. equi bacteria exhaled by clinically healthy and diseased foals were compared using the Mann-Whitney test.
RESULTS
Concentrations of virulent R. equi organisms in exhaled air from foals.Virulent R. equi was detected in exhaled air samples from foals with and without detectable bronchopneumonia. Virulent R. equi was recovered from the breath samples of 37 of the 55 foals tested (67%). The median concentration of virulent R. equi organisms in exhaled air was 8 CFU/1,000 liters. There was no significant difference in the median concentrations of virulent R. equi exhaled by clinically healthy or that of diseased foals (Table 1).
Concentration of virulent R. equi in exhaled air from foals with or without clinical R. equi pneumonia
Concentrations of virulent R. equi in environmental air.All but one farm had detectable airborne virulent R. equi in the lanes and pens during the sampling period. In total, 19/48 (40%) environmental air samples collected across the six farms sampled contained detectable concentrations of virulent R. equi organisms, with a median concentration of virulent R. equi in these positive samples of 2 CFU/1,000 liters. Farms C and F had the greatest number of positive samples (Table 2).
Concentration of airborne virulent R. equi in the pens and lanes on 6 farms during the 2000 foaling season
Comparison of the concentrations of virulent R. equi organisms in environmental and exhaled air samples.The frequency of detection of virulent R. equi in environmental air samples was almost half that of the frequency of detection in exhaled air samples from foals, even though the volume of air sampled from the environment was more than twice that sampled from foals. The median concentration of virulent R. equi organisms in positive samples of exhaled air (10 CFU/1,000 liters) was fivefold higher than that in environmental air samples (2 CFU/1,000 liters). The median concentration of virulent R. equi organisms in all exhaled air samples was significantly higher than the concentration in all environmental air samples (P < 0.001) and was also significantly higher when only positive samples were considered (P < 0.001).
DISCUSSION
Despite being an intracellular pathogen of the equine lower respiratory tract, virulent R. equi was detected in air samples collected near the muzzles of foals. The median concentration of virulent R. equi organisms in breath samples was significantly higher than the median concentration in environmental air samples. These environmental air samples were collected from areas of the farm that have been shown previously to be the most highly contaminated with airborne virulent R. equi (15). This suggested that exhaled air from foals may be a significant source of virulent R. equi for other foals. Studies of experimentally infected foals suggest that the minimum infectious dose of virulent R. equi required to cause pneumonia is approximately 104 CFU/foal, although not all foals inoculated with this dose subsequently developed clinical disease (28). Based on the concentrations of airborne virulent R. equi detected in this study, a foal would require a substantially shorter period of exposure to a contaminated airborne source to acquire this minimum infectious dose when in contact with another foal exhaling high concentrations of virulent R. equi than when only standing and inhaling contaminated dust in a contaminated laneway or holding pen. Foals are unlikely to spend long periods of time in lanes and pens (tens of hours or days) but would routinely spend shorter periods of time (up to several hours) in crowded pens during the breeding season on large horse breeding farms while awaiting veterinary attention or farriery.
Virulent R. equi was detected in exhaled air from 31/48 (65%) of the foals with R. equi pneumonia. In a previous study, Hillidge (8) recovered R. equi from lesions from 7/11 (67%) foals with R. equi pneumonia at necropsy and from 57/89 (64%) tracheal lavage fluid. In a more recent study, R. equi was recovered from 14/21 (67%) transtracheal and 59/96 (61%) nasotracheal lavage fluid from foals with clinical signs of R. equi pneumonia (7). Thus, culture of exhaled air samples appears likely to be at least as sensitive for the detection of clinically affected foals as culture of tracheal lavage specimens, which is still seen as the definitive method for the diagnosis of R. equi pneumonia. Although the sensitivity of air sampling is comparable to that of tracheal lavage fluid samples, the specificity of this test is poor, as 6/7 (86%) healthy foals also exhaled virulent R. equi. These results highlight the difficulties facing clinicians attempting to diagnose R. equi pneumonia, as healthy foals may be a source of infection and so a diagnosis based only on bacteriological detection of virulent organisms in the respiratory tract may be erroneous.
Three apparently healthy foals in this study were found to have lesions suggestive of R. equi pneumonia by ultrasonographic examination. Ultrasonographic examination of lungs has been used widely on horse farms as a diagnostic aid for investigating suspected cases of R. equi pneumonia and as a screening tool on properties with a consistently high prevalence of disease. However, ultrasonographic examination of lungs has limitations in the diagnosis of R. equi pneumonia, as other bacterial causes of pneumonia, such as Streptococcus equi subsp. equi, may induce similar lesions. Furthermore, only superficial lesions are detectable. Deep or mediastinal lesions and small lesions that do not involve the pleural surface of the lung may not be detected (20, 21).
Virulent R. equi was detected in samples of exhaled air from 6/7 clinically healthy foals with no ultrasonographically detectable lung abscesses. This suggests that these foals either had lung abscesses that were undetectable by ultrasonographic examination or were subclinically infected and were shedding virulent R. equi from their respiratory tracts without any obvious lung pathology. As these foals did not develop clinical respiratory illness after sampling, it is more likely that they had subclinical infections with virulent R. equi. Subclinical rhodococcal disease has been seen previously (2). R. equi was recovered from transtracheal lavage fluid from 77/216 (36%) clinically healthy foals in one study (1). That study was performed prior to the differentiation of virulent and avirulent R. equi and hence could not assess the subclinical carriage of virulent R. equi. Subclinical infection may be more common on farms with a high disease prevalence, as these farms are more likely to have elevated environmental burdens of virulent R. equi. The reasons that subclinically infected foals do not develop clinical disease and the potential for these foals to act as respiratory sources of virulent R. equi need to be explored.
We chose to investigate the likelihood of foal-to-foal transmission under field conditions, using quantitative sampling to distinguish the probable contributions of environmental bacteria and those exhaled by foals, because this method was felt to enable a more accurate assessment of the potential significance of the route of infection than one using experimental infection. The ubiquity of R. equi in the environment of horses and suggestions that foals may be infected at a very young age limit the capacity to conduct studies that adequately mimic natural infection and establish contagious transmission under controlled experimental conditions (3, 10). There is no certainty that an intratracheal or intrabronchial bolus of cultured R. equi, the method currently used to experimentally infect foals, would accurately mirror natural infection (12, 28). Such a challenge is likely to result in a distribution of lesions in the respiratory tract that is different from that seen after natural exposure to aerosols over an extended period of time, and thus excretion by experimentally infected foals is not likely to be fully representative of the situation in the field. Virulent R. equi is commonly isolated from the soil on farms and from bedding material and is ingested and excreted in the feces of both diseased and healthy foals, resulting in some degree of environmental contamination on all farms, regardless of the disease status (18, 22, 27). Therefore, there could be no certainty in an experimental model of transmission that the source of infection was the respiratory tract of the infected foals and not organisms excreted in the feces of the foals and then aerosolized from the environment by their movement.
The role of foal-to-foal transmission of virulent R. equi in the epidemiology of R. equi pneumonia requires further investigation. The inhalation of dust contaminated with virulent R. equi is still likely to be an important source of infection, as a high environmental burden of virulent R. equi is correlated with a higher disease prevalence (15, 22). However, foals are often confined in crowded areas such as holding pens for many hours while awaiting veterinary attention and farriery, and even within large paddocks, the likelihood of aerosol transmission from foal to foal may be relatively high. In one study, foals kept in small groups (fewer than five mare-foal pairs per group) were found to have a reduced risk of developing R. equi pneumonia. However, once the year of birth and farm effects were accounted for, the association was not significant (P = 0.071) (4). Even though this finding was not statistically significant, the effect of group size on prevalence warrants further investigation. Management strategies to limit the likelihood of foal-to-foal aerosol transmission, such as decreasing the length of time foals are confined in crowded areas and reducing group sizes, may need to be considered as methods to reduce the prevalence of R. equi pneumonia on farms.
ACKNOWLEDGMENTS
This work was supported by the Rural Industries Research and Development Corporation of Australia.
We thank Garry Anderson for assistance with statistical analysis.
FOOTNOTES
- Received 21 July 2008.
- Returned for modification 12 October 2008.
- Accepted 3 January 2009.
↵▿ Published ahead of print on 14 January 2009.
- American Society for Microbiology
